What are the current incidences and prevalences of HIV by risk factor in the United States? The authors used national HIV surveillance system data collected through 2017 to calculate the number of new diagnoses of HIV in 2015. A previously characterized model of time since infection based on CD4 count was used to estimate the incidence of HIV in 2015, and then this data was paired with information about the cumulative deaths among persons with HIV to generate estimated HIV prevalences. Finally, these data were compared with national estimates of the US populations of men who have sex with men (MSM), people who inject drugs (PWID), and heterosexuals in order to generate population-specific proportions with HIV for each subgroup.
First the good news: new HIV diagnosis rates decreased among all categories between 2010 and 2015 (MSM, 589 to 575 per 100,000 or -2.2%, p=0.012; PWID, 53 to 34 per 100,000 or -35%, p<0.0001; heterosexuals, 5.5 to 4.1 per 100,000 or -25%, p=0.0001). Unfortunately, this trend breaks along racial lines, with black and latinx MSM actually experiencing increasing rates of HIV diagnosis (2191 to 2259 per 100,000 or +3.6% for blacks, p=0.019; 944 to 995 per 100,000 or +5.4% for latinx, p=0.003). Among heterosexuals the improvement in HIV diagnosis rates has been more egalitarian, with decreasing diagnosis rates in all racial, gender and age subgroups except heterosexual men aged 13-24, among whom the HIV diagnosis rate was unchanged.
The current estimated incidences and prevalences of HIV vary by order of magnitudes across risk factors. At present, MSM remains the greatest risk factor for HIV (incidence, 584 per 100,000; prevalence, 124 per 1,000), followed by PWID (incidence, 33 per 100,000; prevalence, 19 per 1,000) and being heterosexual (incidence, 4 per 100,000; prevalence, 1 per 1,000). I take these data to mean that MSM, and to a much lesser extend PWID, must remain the primary focus of HIV treatment and prevention efforts if we intend to end the HIV epidemic. 30585840
CSF in aviremic patients with HIV on ART: does it matter? The literature has gone back and forth on this point, with some studies showing associations between CSF escape and surrogate outcomes such as biomarker of inflammation, but most studies that looked failing to identify any association between CSF viral detection (virrachia? I’m going to call it virrachia) and neurocognitive outcomes. This study examined data from two prospective HIV cohorts, focusing on patients who had “CSF escape,” defined as an undetectable serum viral load on ART plus a CSF with detectable virrachia at the most recent clinic visit. Association with clinical risk factors and outcomes were identified by multivariate analysis.
Of the 1264 patients enrolled in the two cohorts, 55 subjects had CSF escape with a medial viral load of 155 copies/ml. CSF escape was associated with use of protease inhibitors (OR 2.0 for ritonavir-boosted PIs; OR 5.1 for unboosted atazanavir), CSF protein >45 (OR 2.1), and CSF WBC count >5 (OR 7.6). Patients with CSF escape had elevated levels of CSF CXCL10 and lower levels of serum IL-6; however, there were no differences in rates of neurocognitive impairment (38% of both patients who did vs did not have CSF escape had cognitive impairment).
These data agree with previous research on the subject – namely, that CSF escape is associated with use of antiretrovirals with relatively poor CNS penetration, like PIs, and results in increased markers of neuroinflammation but not necessarily any differences in clinically meaningful endpoints. Which begs the question: if my patient has CSF viral escape, should I care? Should I do anything about it? Should I even bother looking for it in the first place? 30702516
Atripla every other day is about as effective as daily dosing for patients who have achieved virologic suppression. Efavirenz has a bad rap for its neuropsychiatric side effects – I’d say not entirely undeserved, based on my experiences with switching a handful of depressed and anxious patients to more modern regimens (though, for the record, retrospective data doesn’t line up with my anecdotal experience, showing no general psychiatric benefit to switching patients off of efavirenz: 30932965). On the bright side, coformulated efavirenz/tenofovir/emtricitabine (which for simplicity’s sake I’ll hereafter call “atripla”) is relatively forgiving of the occasional missed dose.
The authors sought to capitalize on atripla’s pharmacokinetic generosity by comparing the efficacy of daily versus every other day dosing. They performed an open-label RTC of adults with HIV who had achieved viral suppression on atripla for at least six months, randomizing them to either continue daily dosing (n=98) or switch to alternate day dosing (n=99). The primary outcome of interest was viral suppression, defined as <40 copies/ml, at week 48 after randomization.
The cohort was mostly young, (median age 43), male (90%), and MSM (68%); prior to randomization, the median baseline CD4 count was 677 and baseline CD4 nadir was 270. At week 48, rates of virologic suppression were 97% in the daily arm vs 94% in the alternate day arm (risk difference -3%; 95% CI -9% to 3%). One might conclude from this data that the two regimens are equivalent (and indeed the alternate day regimen met the prespecified 10% margin for non-inferiority), but on the other hand you could argue that 10% is too generous for outcome in question, and that by a stricter criterion – say, a margin of 5% - the alternate day schedule failed to show non-inferiority.
But Nico, I hear you saying, wouldn’t it be worth it to give up a trivial amount of efficacy to have patients exposed to less efavirenz and have fewer neuropsychiatric side effects? That’s a great point, except that there were no differences in either the overall rate of serious adverse events or reported neuropsychiatric symptoms; in fact, the event rates were numerically higher in the alternate day therapy arm. So, unless you’re trying to treat your patient’s overwhelming pill fatigue (in which case – hey, good news, the long-acting injectables are on their way!), I don’t see a compelling reason to gamble with alternate day dosing of atripla. 30702517
Whether people think they need PrEP is not a particularly sensitive measure of whether they need PrEP. OK, that may be an overgeneralization – this study was limited to young black MSM in the Southern US, a group who we’ve recently pointed out are at particularly high risk of HIV transmission. But people are people and I suspect the point would hold for other populations, even if the magnitude of additional risk is lower. The authors interviewed 29 subjects about their perceived need for PrEP, focusing on reasons the subjects accepted or refused PrEP and why their decisions might change in the future. This study is an element of the ELEMENT study, a prospective cohort examining substance use and HIV incidence among young black MSM in Atlanta over a 2-year period; the subjects were recruited based on peer referral and advertising about men’s health and HIV that appeared on Grindr (a dating/hookup app primarily aimed at MSM), Facebook, and local transit systems.
The 29 subjects interviewed were mostly educated (72% had completed some college or more) and poor (38% had an annual income under $40,000). Just under half had been prescribed PrEP, but only 38% of these were taking it consistently; of those not prescribed PrEP, the majority indicated they were undecided about PrEP or might consider it under other circumstances. When examining reasons for not using PrEP further, the authors found subjects’ self-perceived need for PrEP was low, with many citing being in a monogamous relationship, consistently using condoms, or simply not being at risk for HIV transmission (e.g. because they were not sexually active at the time of interview) as reasons for not needing PrEP. Many subjects also viewed sexually activity as being on a scale, with language like “more active” indicating that, while engaging in sex, the amount of sex they were having did not pose enough risk to require PrEP; others did not perceive themselves to be “sexually active” unless they were having frequent encounters with multiple partners. Finally, several subjects reported prior STI diagnoses but failed to connect these to a need for PrEP.
Follow-up surveys of this cohort demonstrated that the participants had consistently been overly optimistic about their need for PrEP. Among those who had deferred PrEP due to not being sexually active, 75% subsequently reported having had anal sex in the interim. Among those who deferred PrEP due to being in a monogamous relationship, 50% later disclosed that their relationships were not mutually monogamous and/or that they did not know their partner’s HIV status. Among those who cited consistent condom use as the reason they did not need PrEP, 69% reported having had unprotected anal sex in a follow-up survey.
So, to summarize: people are ignorant about the risk factors indicating PrEP, overly optimistic about their current level of risk of HIV transmission, and overly optimistic about their future use of safe sex behaviors (e.g. using condoms). This is why, IMO, the PrEP guidelines should be condensed to “if you [have sex], take PrEP.” 30844305
Finally, a brief update on HIV cure research: a new toll-like receptor 9 agonist therapy may help patients maintain viral suppression in the absence of ART. That’s a bold claim based on data from one of nine patients responding to said therapy, but as a pilot study I think this is interesting. The clinical trial enrolled twelve patients on suppressive ART, treating them with 24 weeks of the TLR9 agonist lefitolimod. The treatment was safe and improved anti-HIV innate immune responses (e.g. HIV-specific T-cell activation), though it did not reduce patients’ HIV DNA / proviral reservoir. While there was no benefit to time to virologic rebound among nine patients randomized to stop ART, one of the patients was able to control their viremia without ART for five months. While that could be a fluke, more TLR9 agonist therapy research is clearly indicated. 30932955